Geopolitics & Sovereign Positioning

The US military is operationally using AI systems from Palantir and Anthropic to transform battlefield intelligence into strike decisions against Iranian targets, according to Financial Times. The systems process torrents of sensor data, satellite imagery, and signals intelligence to generate targeting recommendations that are then executed by human operators. This represents the first sustained use of what military planners call an AI-driven kill chain in active combat. Separately, a pro-Iran hacking group claimed responsibility for a crippling cyberattack on US medical technology firm Stryker, which remains offline with no timeline for restoration, according to Bloomberg. Neither Stryker nor US cyber agencies have confirmed Iranian state involvement, but the attack's sophistication and timing suggest coordination with Tehran's broader asymmetric response to US strikes.

Meanwhile, offensive cyber capabilities in frontier AI models are advancing rapidly, with models progressing from near-zero to meaningful success rates on expert-level security challenges within months, according to research from IAPS. Leading AI developers are now triggering their own internal risk thresholds for cybersecurity capabilities, and real-world cases have emerged where AI agents autonomously executed significant portions of state-sponsored cyber campaigns. The convergence of AI-enabled offensive cyber and autonomous targeting systems is fundamentally altering strategic stability calculations.

China is establishing a network of humanoid robot training facilities to generate the data needed to develop autonomous machine intelligence, according to Financial Times. These training farms aim to supply the massive datasets required to put functional AI brains into physical robots, addressing a key bottleneck in China's push for robotics leadership. Simultaneously, Chinese adoption of OpenClaw, an open-source autonomous AI agent developed by an Austrian programmer, has exploded into a gold rush, with entrepreneurs rapidly building businesses around the tool that can control devices and autonomously complete complex tasks, according to MIT Technology Review. Beijing is funding this technology infrastructure push by extracting unprecedented profit clawbacks from state-owned enterprises, drawing on them more than ever as government budget revenue saw an almost unprecedented decline, according to Bloomberg.

The coordinated push across robotics training data, autonomous agent adoption, and financial reallocation signals Beijing's determination to build indigenous AI capacity despite export controls. China is treating AI development as a national mobilisation challenge, squeezing domestic resources to reduce dependency on Western technology while racing to establish leadership in embodied AI and agentic systems before US controls tighten further.

At least 11 African governments have spent more than $2 billion on Chinese-built AI-powered mass surveillance systems featuring facial recognition and behavioural tracking capabilities, according to The Guardian. Human rights experts warn these systems are neither necessary nor proportionate, violate citizens' privacy rights, and create chilling effects on civil society. The surveillance infrastructure creates lasting dependencies on Chinese technology providers for maintenance, upgrades, and operator training, while generating vast datasets on African populations that flow to Chinese companies. This represents Beijing's most successful technology export in creating durable influence relationships, as recipient governments become locked into Chinese ecosystems and reliant on continued access to system operators and spare parts.

Nvidia will spend $26 billion to build open-weight AI models, according to WIRED, positioning the chip giant to compete directly with OpenAI, Anthropic, and DeepSeek across the full AI stack from infrastructure to applications. Simultaneously, Nvidia struck a $2 billion deal with AI cloud provider Nebius, according to Financial Times, continuing a dealmaking spree that deploys its massive cash reserves into funding its own customers' infrastructure purchases. The combined moves represent Nvidia's strategy to capture value across multiple layers of the AI value chain while ensuring continued demand for its core chip business by financing customers' expansion and building complementary software that runs on Nvidia hardware.

The UK government will remove commercially available AI systems from mandatory national security investment screening as part of a deregulation drive to reduce red tape, while simultaneously requiring all water company takeovers to undergo government security review, according to Financial Times. The policy reveals fragmented threat assessment where physical infrastructure receives heightened protection while software systems that increasingly control critical functions face reduced scrutiny. The move comes as the UK struggles to attract the tens of billions in AI infrastructure investment the government has promised, with recent investigations revealing discrepancies between announced datacenter projects and actual construction plans.